[Bug-gnubg] Question on Luck Analysis

[Ian Shaw]

This query is prompted by a discussion arising during the Variance
Reduction Tournament taking place at Daily Gammon.

Below are the results from a single cubeless game where gnubg 0-ply
played itself, with a 0-ply analysis. This means that both sides are
judged to have played perfectly, and without a cube there can be no
discrepancies in luck caused by the cube.

In a game such as this, how can one side win a game without having +0.5
points of net luck? If neither side makes an error, then the only way to
get from 50% to 100% is to have 50% luck. 

However, Blue has +0.270 luck and White had -.863 luck, giving Blue net
luck of +1.133 points, more than needed to win the game.

If both sides are perfect, shouldn't the luck-adjusted result be 0?

Gnubg is clearly reporting something other than what I expect. Can
anybody explain it?


                                      Blue                 White

Chequer Play Statistics:
Total moves                            26                   25

Unforced moves                         26                   21

Unmarked moves                         26                   25

Moves marked doubtful                   0                    0

Moves marked bad                        0                    0

Moves marked very bad                   0                    0

Error total EMG (Points)                   +0.000 ( +0.000)      +0.000
( +0.000)
Error rate mEMG (Points)                   +0.0   ( +0.000)      +0.0
( +0.000)
Chequerplay rating                    Supernatural         Supernatural


Luck Statistics:
Rolls marked very lucky                 0                    0

Rolls marked lucky                      0                    0

Rolls unmarked                         26                   24

Rolls marked unlucky                    0                    1

Rolls marked very unlucky               0                    0

Luck total EMG (Points)                    +0.270 ( +0.270)      -0.863
( -0.863)
Luck rate mEMG (Points)                   +10.4   ( +0.010)     -34.5
( -0.035)
Luck rating                           None                 Bad dice,
man!      

Overall Statistics:
Error total EMG (Points)                   +0.000 ( +0.000)      +0.000
( +0.000)
Error rate mEMG (Points)                   +0.0   ( +0.000)      +0.0
( +0.000)
Snowie error rate                          +0.0   ( +0.000)      +0.0
( +0.000)
Overall rating                        Supernatural         Supernatural

Actual result                          +1.000               -1.000

Luck adjusted result                   -0.133               +0.133



Cheers,
Ian Shaw